This script leverages the power of OpenAI's GPT-3.5 Turbo to automate the academic article screening process. It processes an XML file of articles exported from EndNote, evaluates their relevance based on PICOS criteria, and saves the results in both CSV and XML formats.

To install the required Python packages, run the following command:

pip install -r requirements.txt

• beautifulsoup4
• pandas
• openai
• tqdm
• tenacity

Here are the command-line arguments that the script accepts:

Here's how you can run the script:

python script.py --xml_file "path/to/xml/file.xml" --systemprompt "You are a medical researcher AI." --preprompt "You are an AI assistant..." --prompt "You are given an article..." --postprompt "Population (P): Patients..." --output "path/to/output/file" --apikey "YOUR_API_KEY" --model "gpt-3.5-turbo" --temperature 0.8 --interval 20.0 --sleep "true"

The script generates two output files:

• A new XML file with the ratings added.
• A CSV file with the article details, ratings, and AI responses.

• Make sure to replace "YOUR_API_KEY" with your actual OpenAI API key.
• The script employs exponential backoff to handle API rate limits or failures gracefully.

If you use this script in your research or project, please cite it using the following: